Accuracy and reliability of a subcutaneous continuous glucose monitoring device in critically ill patients.

Subcutaneous continuous glucose monitoring (CGM) may have benefits in achieving glycemic control in critically ill patients. The aim of this study was to assess the accuracy and reliability of the FreeStyle Navigator I in critically ill patients and to assess patient related factors influencing the accuracy and reliability. This study is a retrospective analysis of data from a randomized controlled trial conducted in a 20-bed mixed intensive care unit. Analytical accuracy, clinical accuracy and reliability were assessed against arterial blood glucose samples as reference. Assessment was according to recent consensus recommendations with median absolute relative difference (median ARD), Bland-Altman plots, the ISO system accuracy standards (ISO 15197:2013) and Clarke error grid analysis (CEG). We analyzed 2840 paired measurements from 155 critically ill patients. The median ARD of all paired values was 13.3 [6.9-22.1]%. The median ARD was significantly higher in both the hypoglycemic and the hyperglycemic range (32.4 [12.1-53.4]% and 18.7 [10.7-28.3]% respectively, p < 0.001). The Bland-Altman analysis showed a mean bias of - 0.82 mmol/L with a lower limit of agreement (LOA) of - 3.88 mmol/L and an upper LOA of 2.24 mmol/L. A total of 1626 (57.3%) values met the ISO-2013, standards and 1,334 (47%) CGM values were within 12.5% from the reference value. CEG: 71.0% zone A, 25.8% zone B, 0.5% zone C, 2.5% zone D, 0.3% zone E. The median overall real-time data display time was 94.0 ± 14.9% and in 23% of the patients, the sensor measured < 95% of the time. Additionally, data gaps longer than 30 min were found in 48% of the patients. The analytical accuracy of the FreeStyle Navigator I in critically ill patients was suboptimal. Furthermore, the clinical accuracy, did not meet the required standards. The reliability was satisfactory, however, in almost a quarter of the patients the realtime data display was < 95%. The accuracy was considerably and significantly lower in hyper- and hypoglycemic ranges.

Pain measurement in mechanically ventilated critically ill patients: Behavioral Pain Scale versus Critical-Care Pain Observation Tool.

PURPOSE: The Behavioral Pain Scale (BPS) and Critical-Care Pain Observation Tool (CPOT) are behavioral pain assessment tools for uncommunicative and sedated intensive care unit (ICU) patients. This study compares the discriminant validation and reliability of the CPOT and the BPS, simultaneously, in mechanically ventilated patients on a mixed-adult ICU. MATERIALS AND METHODS: This is a prospective observational cohort study in 68 mechanically ventilated medical ICU patients who were unable to report pain. RESULTS: The BPS and CPOT scores showed a significant increase of 2 points between rest and the painful procedure (turning). The median BPS scores between rest and the nonpainful procedure (oral care) showed a significant increase of 1 point, whereas the median CPOT score remained unchanged. The interrater reliability of the BPS and CPOT scores showed a fair to good agreement (0.74 and 0.75, respectively). CONCLUSIONS: This study showed that the BPS and the CPOT are reliable and valid for use in a daily clinical setting. Although both scores increased with a presumed painful stimulus, the discriminant validation of the BPS use was less supported because it increased during a nonpainful stimulus. The CPOT appears preferable in this particular group of patients, especially with regard to its discriminant validation.